The honest answer is: it depends entirely on the product in question, and the gap between the best and worst-evidenced options on the market is far wider than marketing from both sides would suggest. Here's what the evidence actually shows, category by category.
Melatonin: truly effective, but only for a specific problem
Melatonin has one of the larger evidence bases of all sleep supplements, but its effectiveness is specific to circadian timing problems, jet lag, shift work, and a shifted sleep window, not for insomnia in general. Clinical research supports doses in the range of 0.5 to 3 milligrams, significantly less than many commercial products now contain, and higher doses have not shown significantly better results. For someone whose sleep problem is truly circadian, the evidence is solid. For someone whose problem is stress-induced or respiratory, melatonin addresses a mechanism that was never the cause.
Magnesium: moderate support, often exaggerated
For magnesium, there is clinical evidence for improved sleep quality, especially in people with magnesium deficiency and in older adults. The presumed mechanism is via its role in regulating GABA, a calming neurotransmitter. The effect sizes in the available studies are generally moderate, and much of the more dramatic marketing around magnesium in the current wellness market outpaces the actual research.
Valerian and herbal blends: weak, inconsistent evidence
Valerian is one of the most commonly used herbal sleep aids, and also one of the most inconsistently evidenced. Systematic reviews of valerian studies find mixed results; some studies show moderate benefit, others no significant difference from placebo. Researchers usually attribute this pattern to inconsistent extraction methods and dosages of different products; there is no definitive answer as to whether the plant itself works.
CBD: promising, but very early
Research into CBD for sleep is growing rapidly, but clinical evidence in humans remains comparatively limited. Most existing studies have small sample sizes or measure anxiety rather than sleep directly. Initial research suggests a potential benefit, especially for sleep disturbances related to anxiety, but sleep researchers remain generally cautious about treating CBD as an established sleep measure, because current market euphoria has significantly outpaced the size and quality of the evidence base.
Weighted blankets: moderate, real, but limited evidence
Several smaller studies show that weighted blankets can improve subjective sleep quality and shorten the time it takes to fall asleep. The presumed mechanism is through deep pressure stimulation and its effect on the nervous system. The evidence is real but limited in scope, and the effect appears to be more pronounced in anxiety-related sleep problems than in sleep disorders with other causes.
Sleep trackers: useful for patterns, not for diagnoses
Consumer sleep trackers have improved significantly, but validation studies compared to clinical polysomnography continue to find relevant inaccuracies, especially in precisely distinguishing sleep stages. Their true value lies in recognizing patterns over time – a regular drop in sleep quality in a specific week of the cycle, a consistent disturbance after alcohol – not in a clinically exact night-by-night evaluation.
Nasal strips and airway aids: real mechanism, evidence for broad application still developing
The mechanism behind nasal strips, supporting airflow through a mechanically constricted nasal passage, is well-documented in respiratory physiology. Less clear is how much someone without an apparent constriction benefits from nightly use. The honest position, which sleep physicians also hold regarding this category, is: the tool is most clearly useful for people with an identifiable limitation of nasal airflow, not as a universal night-time add-on with guaranteed benefits for everyone.
Where NMN stands in this evaluation
NMN falls into a different category than most of the above because its goal is not sleep directly, but NAD+, the coenzyme that activates SIRT1, the protein behind the precision of the circadian clock. The clinical study situation is smaller than for melatonin, but mechanistically more specific than for most herbal options: a randomized, double-blind, placebo-controlled study from 2022 in the journal Nutrients found significant improvements in sleep quality scores in adults who took NMN daily for twelve weeks. As with most categories above: the research base is real, but compared to more established measures, it is still developing, and any claim that goes beyond what the published studies themselves report deserves the same skepticism as in any other category on this list.
The honest evaluation framework for all of this
The products with the strongest evidence share three characteristics: a concrete, identified mechanism, a dose that corresponds to what has actually been studied, and a defined target group in which the research took place. A product that promises broad, dramatic effects for everyone, at a dose or formulation that has never been directly tested, is claiming something that the evidence does not support, regardless of the category. The question worth asking is never simply "does it work?". It is "for which specific problem, in which target group, at what dose has this actually been shown?", because that is precisely the question most sleep marketing tries to avoid.